KR102582291B1 - Emotion information-based voice synthesis method and device - Google Patents

Abstract

Disclosed is a voice synthesis method and device based on emotion information. In the emotion information-based voice synthesis method according to an embodiment of the present invention, when emotion information is set in a received message, metadata corresponding to the set emotion information is transmitted to the voice synthesis engine, and the emotion information is included in the received message. If not set, new emotional information is generated through semantic analysis and context analysis within the sentences of the received message, and metadata corresponding to the generated emotional information is transmitted to the speech synthesis engine. The voice synthesis engine can perform voice synthesis by loading emotional information based on the transmitted metadata.

Description

Emotion information-based voice synthesis method and device

The present invention relates to a voice synthesis device and method, and more specifically, to a voice synthesis device and method that allows output of voice with emotional content.

Conventional text-to-speech (TTS) processing outputs text using pre-stored speech. The primary purpose of TTS processing is to convey the semantic contents of the text, but recently, TTS processing allows not only the semantic contents of the text but also the interactive meaning of the text to be conveyed to the other party. , there is a need to ensure that the intention or emotions of the user who delivered the actual text are reflected in the voice output so that the user can actually experience an interactive conversation with the text sender.

The present invention aims to solve the above-described needs and/or problems.

Additionally, the present invention aims to implement voice synthesis that outputs a voice that reflects the transmitter's intent.

Additionally, the present invention aims to implement voice synthesis containing emotions.

Additionally, the present invention aims to implement voice synthesis containing emotions by analyzing the semantic content and context information of the message to be conveyed.

An emotional information-based voice synthesis method according to an aspect of the present invention includes receiving a message; determining whether emotional information is set in the received message; When first emotion information is set in the message, transmitting first metadata corresponding to the first emotion information to a voice synthesis engine; When emotion information is not set in the message, generate second emotion information based on semantic analysis and context analysis of the message, and transmit second metadata corresponding to the second emotion information to the speech synthesis engine. step; and, by the voice synthesis engine, synthesizing a voice corresponding to the message by adding emotional information determined based on either the first metadata or the second metadata.

The first emotion information may be set separately from the input of the message.

The first metadata and the second metadata are described in a markup language format, and the markup language may include Speech Synthesis Markup Language (SSML).

The SSML includes elements representing emotional properties, and the emotional properties include neutral, love, happy, angry, sad, worry, or sorry. ) may include at least one of

Generating the second emotional information includes transmitting the message to a semantic analysis module and a context analysis module; The semantic analysis module includes: transmitting a first emotion vector calculated based on emotional elements included in the message from which emotions can be inferred to an emotional information determination module; And the context analysis module may include transmitting, by the context analysis module, a second emotion vector calculated based on the entire context of the message to the emotion information determination module.

The emotion information-based voice synthesis method includes: the emotion information determination module determining the second emotion information to be reflected in speech synthesis based on the first emotion vector and the second emotion vector; and generating the second metadata corresponding to the second emotion information and transmitting it to the speech synthesis engine. It may further include.

The first emotion vector may be defined as a weight given to a plurality of emotional attributes by normalizing the sum of the weights, and the second emotion vector may be defined as a weight given to the plurality of emotional attributes by normalizing the sum of the weights. You can.

Determining the second emotion information may include adding a first emotion vector to which a first weight is applied and a second emotion vector to which a second weight is applied; and selecting, as the second emotional information, an emotional attribute with the largest value as a result of the summation.

The first weight and the second weight may be adaptively changed according to a result of analyzing the meaning and context of the message, and increasing the first weight if the message contains emotional expressions; and increasing the second weight when it is determined that there is continuity between the plurality of sentences constituting the message. However, the first weight and the second weight may be normalized and defined.

The first weight and the second weight may be recorded and stored for each sender who sent the message.

The weights assigned to the plurality of emotional attributes constituting the first emotional vector may be assigned in consideration of symbols or graphic objects included in the message as a result of inferring semantic contents included in the message. You can.

The weights assigned to the plurality of emotional attributes constituting the second emotional vector may be given by considering the context between sentences from which the flow of context can be inferred.

The emotional element that can infer the emotion may be defined as at least one of a text, symbol, or graphic object included in the received message.

A voice synthesis method based on emotional information according to another aspect of the present invention includes receiving a message from a transmitting device; And when it is determined that emotional information is set in the message, defining metadata corresponding to the emotional information in SSML (Speech Synthesis Markup Language) format and transmitting it to the speech synthesis device along with the message.

A voice synthesis device based on emotion information according to another aspect of the present invention includes a communication unit that receives a message; a voice synthesis unit that synthesizes voice corresponding to the message; and a processor that controls the voice synthesis unit to perform a voice synthesis operation based on emotion information based on whether emotion information is set in the received message, wherein the processor is configured to include first emotion information in the message. When set, first metadata corresponding to the first emotion information is transmitted to the voice synthesis unit, and when emotion information is not set in the message, the emotion element that can infer the emotion included in the message or the Generates second emotion information based on at least one of the context of the message and transmits second metadata corresponding to the second emotion information to the voice synthesis unit, and the voice synthesis unit generates the first metadata or the second metadata. Emotion information determined based on one of the metadata is added to synthesize a voice corresponding to the message.

A voice synthesis system based on emotional information according to another aspect of the present invention includes a transmission device for transmitting a message; a server that defines metadata corresponding to emotional information extracted through a semantic analysis process and a context analysis process for the message received from the transmitting device in SSML (Speech Synthesis Markup Language) format and transmits it together with the message; and a voice synthesis device that synthesizes a voice reflecting the emotional information based on the metadata.

A computing device according to another aspect of the present invention includes a processor and a memory including instructions executable by the processor, wherein the instructions determine whether emotional information is set in the received message, When first emotion information is set in the message, the first metadata corresponding to the first emotion information is transmitted to the speech synthesis engine, and when emotion information is not set in the message, it is included in the message to infer the emotion. Generating second emotion information based on at least one of possible emotion elements or the context of the message, and transmitting second metadata corresponding to the second emotion information to a speech synthesis engine, wherein the speech synthesis engine , emotion information determined based on either the first metadata or the second metadata is added to synthesize a voice corresponding to the message.

The effects of the emotion information-based voice synthesis method and voice synthesis device according to the present invention will be described as follows.

The present invention can implement voice synthesis that outputs a voice that reflects the transmitter's intent.

Additionally, the present invention can implement voice synthesis containing emotions.

Additionally, the present invention can output a message containing emotions for each sentence by analyzing the semantic content and context information of the message to be conveyed.

Further scope of applicability of the present invention will become apparent from the detailed description that follows. However, since various changes and modifications within the spirit and scope of the present invention may be clearly understood by those skilled in the art, the detailed description and specific embodiments, such as preferred embodiments of the present invention, should be understood as being given by way of example only.

1 is a block diagram of a communication system according to an embodiment of the present invention.
Figure 2 is a block diagram of a communication system according to another embodiment of the present invention.
Figure 3 illustrates communication between two users via part of the communication system of Figure 1 of the present invention.
Figure 4 shows a schematic block diagram of a voice synthesis device in a voice synthesis system environment according to an embodiment of the present invention.
Figure 5 shows a schematic block diagram of a voice synthesis device in a voice synthesis system environment according to another embodiment of the present invention.
Figure 6 shows a schematic block diagram of an artificial intelligence agent capable of implementing voice synthesis based on emotion information according to an embodiment of the present invention.
Figure 7 is another block diagram of a voice synthesis device based on emotion information according to an embodiment of the present invention.
Figure 8 is a flowchart of a voice synthesis method based on emotion information according to an embodiment of the present invention.
Figure 9 is a flowchart of a voice synthesis method based on emotion information for generating new emotion information when emotion information is not set in a received message according to an embodiment of the present invention.
Figure 10 is a diagram for explaining an emotion vector according to an embodiment of the present invention.
Figure 11 shows an example of calculating a first emotion vector by analyzing the meaning of the message when emotion information is not set in the received message according to an embodiment of the present invention.
Figure 12 shows an example of calculating a second emotion vector by analyzing the context of the message when emotion information is not set in the received message according to an embodiment of the present invention.
Figure 13 shows an example of determining final emotion information based on the first emotion vector and the second emotion vector when emotion information is not set in the received message according to an embodiment of the present invention.
Figure 14 is a flowchart of a voice synthesis method based on emotion information according to an embodiment of the present invention.
Figure 15 shows an example in which emotional information is described in SSML language format according to an embodiment of the present invention.
Figure 16 is a flowchart for explaining the operation of a voice synthesis system according to an embodiment of the present invention.
Figures 17a and 17b show an example of sending a message by setting emotional information according to an embodiment of the present invention.

Hereinafter, embodiments disclosed in the present specification will be described in detail with reference to the attached drawings. However, identical or similar components will be assigned the same reference numbers regardless of reference numerals, and duplicate descriptions thereof will be omitted. The suffixes “module” and “part” for components used in the following description are given or used interchangeably only for the ease of preparing the specification, and do not have distinct meanings or roles in themselves. Additionally, in describing the embodiments disclosed in this specification, if it is determined that detailed descriptions of related known technologies may obscure the gist of the embodiments disclosed in this specification, the detailed descriptions will be omitted. In addition, the attached drawings are only for easy understanding of the embodiments disclosed in this specification, and the technical idea disclosed in this specification is not limited by the attached drawings, and all changes included in the spirit and technical scope of the present invention are not limited. , should be understood to include equivalents or substitutes.

Terms containing ordinal numbers, such as first, second, etc., may be used to describe various components, but the components are not limited by the terms. The above terms are used only for the purpose of distinguishing one component from another.

When a component is said to be "connected" or "connected" to another component, it is understood that it may be directly connected to or connected to the other component, but that other components may exist in between. It should be. On the other hand, when it is mentioned that a component is “directly connected” or “directly connected” to another component, it should be understood that there are no other components in between.

Singular expressions include plural expressions unless the context clearly dictates otherwise.

In this application, terms such as “comprise” or “have” are intended to designate the presence of features, numbers, steps, operations, components, parts, or combinations thereof described in the specification, but are not intended to indicate the presence of one or more other features. It should be understood that this does not exclude in advance the possibility of the existence or addition of elements, numbers, steps, operations, components, parts, or combinations thereof.

1 is a block diagram of a communication system according to a preferred embodiment of the present invention.

Referring to FIG. 1, the communication system includes at least one transmitting device 12, at least one receiving device 14, and at least one network system 16 connecting the transmitting device 12 with the receiving device 14. ), and may include a Text-To-Speech (TTS) system 18 as a speech synthesis engine.

The at least one transmitting device 12 and the at least one receiving device 14 include mobile phones 21 and 31, smart phones, digital broadcasting terminals, personal digital assistants (PDAs), and portable multimedia players (PMPs). , navigation, ultrabook, wearable device (e.g., smartwatch), glass-type terminal (smart glass), head mounted display (HMD), etc. may be included.

The at least one transmitting device 12 and the at least one receiving device 14 include a slate PC (22, 32), a tablet PC, a laptop computer (23, 33), etc. It may further include. The PCs 22 and 32 and laptop computers 23 and 33 can be connected to the at least one network system 16 through wireless access points 25 and 35.

The at least one transmitting device 12 and the at least one receiving device 14 may be called a client device.

Figure 2 is a block diagram of a communication system according to another embodiment of the present invention. The example of the communication system shown in Figure 2 is similar to the communication system shown in Figure 1 except for the TTS system. The omitted TTS system may be included in at least one transmitting device 12. That is, unlike the environment in FIG. 1, FIG. 2 shows that the TTS system can be implemented inside the transmitting device 12 so that the TTS system can be implemented with on-device processing.

1 and 2 show exemplary embodiments of the invention. 1 and 2 are intended to provide a context in which the features of the present invention can be implemented. More specific descriptions of one or more system architectures implementing such systems may be provided elsewhere herein, through integrated applications, etc. In addition, in terms of text messaging, each of the communication systems shown in FIGS. 1 and 2 preferably includes a communication network, and in the context of instant messaging, the communication shown in FIGS. 1 and 2 Systems 10 preferably include the Internet.

Figure 3 illustrates communication between two users U1 and U2 via part of the communication system of Figure 1 of the present invention.

Referring to FIG. 3, a first user (U1), which may be referred to as a transmitting user, may be referred to as a receiving user through the transmitting device 12 and the receiving device 14. Can communicate with user (U2). The first user (U1) can use his/her sending device (12) to initiate a text message (Ex, I want to see you. Where are you?^^) which is sent to the second user (U2) via the receiving device (14). there is. Additionally, in terms of text messaging, a first user 32 may send an instant message via an Instant Messaging (IM) client, and a second user 34 may send an instant message on the receiving device 14 via an IM client. You can receive instant messages.

The message sent by the first user (U1) may be transmitted to the TTS system 18. The speech synthesis engine of the TTS system 18 can convert the message into speech. The second user (U2) can hear the synthesized voice corresponding to the received message from the TTS system 180 being output through the receiving device 14. The transmitting device 12 and receiving device 14 may include a microphone, speaker, and display.

Hereinafter, the voice processing process performed in a device environment and/or a cloud environment or server environment will be described with reference to FIGS. 4 and 5. FIG. 4 illustrates an example in which voice input can be performed in the device 50, but the process of processing the input voice and synthesizing voice, that is, the overall operation of voice processing, is performed in the cloud environment 60. On the other hand, FIG. 5 shows an example of on-device processing in which the overall voice processing operation of processing the above-mentioned input voice and synthesizing voice is performed in the device 70.

4 and 5, the device environments 50 and 70 may be referred to as client devices, and the cloud environments 60 and 80 may be referred to as servers.

Figure 4 shows a schematic block diagram of a voice synthesis device in a voice synthesis system environment according to an embodiment of the present invention.

Various components are required to process voice events in an end-to-end voice UI environment. The sequence for processing speech events includes signal acquisition and playback, speech pre-processing, voice activation, speech recognition, natural language processing, and Finally, the device performs a speech synthesis process to respond to the user.

Client device 50 may include an input module. The input module may receive user input from the user. For example, the input module can receive user input from a connected external device (eg, keyboard, headset). Also, for example, the input module may include a touch screen. Also, for example, the input module may include hardware keys located on the user terminal.

According to one embodiment, the input module may include at least one microphone capable of receiving a user's speech as a voice signal. The input module includes a speech input system and can receive the user's speech as a voice signal through the speech input system. The at least one microphone may determine a digital input signal for the user's speech by generating an input signal for audio input. According to one embodiment, a plurality of microphones may be implemented as an array. The array may be arranged in a geometric pattern, such as linear geometry, circular geometry, or any other configuration. For example, for a given point, an array of four sensors may be arranged in a circular pattern separated by 90 degrees to receive sound from four directions. In some implementations, the microphone may include a spatially disparate array of sensors in data communication, which may include a networked array of sensors. Microphones may include omnidirectional, directional (eg, shotgun microphone), etc.

The client device 50 may include a pre-processing module 51 that can preprocess a user input (voice signal) received through the input module (eg, microphone).

The preprocessing module 51 includes an adaptive echo canceller (AEC) function, thereby removing echoes included in the user's voice signal input through the microphone. The preprocessing module 51 can remove background noise included in user input by including a noise suppression (NS) function. The pre-processing module 51 includes an end-point detect (EPD) function, so that it can detect the end point of the user's voice and find the part where the user's voice exists. In addition, the preprocessing module 51 includes an automatic gain control (AGC) function, so that it can recognize the user input and adjust the volume of the user input to be suitable for processing.

Client device 50 may include a voice activation module 52. The voice recognition activation module 52 can recognize a wake up command to recognize a user's call. The voice recognition activation module 52 can detect a predetermined keyword (ex, Hi LG) from user input that has gone through a preprocessing process. The voice recognition activation module 52 is in a standby state and can perform an always-on keyword detection function.

The client device 50 may transmit the user's voice input to the cloud server. Automatic speech recognition (ASR) and natural language understanding (NLU), which are core components for processing user voice, are traditionally run in the cloud due to computing, storage, and power constraints. The cloud may include a cloud device 60 that processes user input transmitted from a client. The cloud device 60 may exist in the form of a server.

The cloud device 60 includes an Auto Speech Recognition (ASR) module 61, an Artificial Intelligent Agent (62), a Natural Language Understanding (NLU) module 63, and text-to-speech conversion ( It may include a Text-to-Speech (TTS) module 64 and a service manager 65.

The ASR module 61 may convert the user voice input received from the client device 50 into text data.

The ASR module 61 includes a front-end speech pre-processor. A front-end speech preprocessor extracts representative features from speech input. For example, a front-end speech preprocessor performs a Fourier transform on the speech input to extract spectral features that characterize the speech input as a sequence of representative multidimensional vectors. Additionally, the ASR module 61 may include one or more speech recognition models (eg, acoustic models and/or language models) and implement one or more speech recognition engines. Examples of speech recognition models include hidden Markov models, Gaussian-Mixture Models, Deep Neural Network Models, n-gram language models, and other statistical models. Examples of speech recognition engines include dynamic time warp based engines and weighted finite state transformer (WFST) based engines. One or more speech recognition models and one or more speech recognition engines produce intermediate recognition results (e.g., phonemes, phoneme strings, and subwords) and ultimately text recognition results (e.g., words, word strings, or tokens). It can be used to process the extracted representative features in a front-end speech preprocessor to generate a sequence.

Once the ASR module 61 generates a recognition result that includes a text string (e.g., words, or a sequence of words, or a sequence of tokens), the recognition result is sent to the natural language processing module 732 for intent inference. It is delivered. In some examples, ASR module 730 generates multiple candidate text representations of the speech input. Each candidate text representation is a sequence of words or tokens corresponding to speech input.

The NLU module 63 can determine user intent by performing syntactic analysis or semantic analysis. The grammatical analysis divides grammatical units (eg, words, phrases, morphemes, etc.) and determines what grammatical elements the divided units have. The semantic analysis can be performed using semantic matching, rule matching, formula matching, etc. Accordingly, the NUL module 63 can obtain the domain, intent, or parameters necessary for the user input to express the intent.

The NLU module 63 can determine the user's intention and parameters using a mapping rule divided into domain, intention, and parameters necessary to identify the intention. For example, one domain (e.g., alarm) may include multiple intents (e.g., set alarm, disarm alarm), and one intent may include multiple parameters (e.g., time, repeat number of times, alarm sound, etc.). A plurality of rules may include, for example, one or more required element parameters. The matching rules may be stored in a Natural Language Understanding Database.

The NLU module 63 uses linguistic features (e.g., grammatical elements) such as morphemes and phrases to identify the meaning of words extracted from user input, and matches the meaning of the identified word to the domain and intent. determine the user's intention. For example, the NLU module 63 may determine the user intent by calculating how many words extracted from the user input are included in each domain and intent. According to one embodiment, the NLU module 63 may determine parameters of user input using words that are the basis for identifying the intent. According to one embodiment, the NLU module 63 may determine the user's intention using a natural language recognition database in which linguistic features are stored to determine the intention of the user input. Also, according to one embodiment, the NLU module 63 may determine the user's intention using a personal language model (PLM). For example, the NLU module 63 may determine the user's intention using personalized information (eg, contact list, music list, schedule information, social network information, etc.). The personalized language model may be stored, for example, in a natural language recognition database. According to one embodiment, not only the NLU module 63 but also the ASR module 61 may recognize a user's voice by referring to a personalized language model stored in a natural language recognition database.

The NLU module 63 may further include a natural language generation module (not shown). The natural language generation module can change specified information into text form. The information changed to the text form may be in the form of natural language speech. The designated information may include, for example, information about additional input, information guiding the completion of an operation corresponding to a user input, or information guiding the user's additional input. The information changed into text form may be transmitted to a client device and displayed on a display, or transmitted to a TTS module and changed into voice form.

The speech synthesis module (TTS module, 64) can change information in text form into information in voice form. The TTS module 64 may receive information in the form of text from the natural language generation module of the NLU module 63, change the information in the form of text into information in the form of voice, and transmit it to the client device 50. The client device 50 may output the information in the form of voice through a speaker.

Speech synthesis module 64 synthesizes speech output based on the provided text. For example, the result generated by the speech recognition module (ASR) 61 is in the form of a text string. Speech synthesis module 64 converts the text string into audible speech output. Speech synthesis module 64 uses any suitable speech synthesis technique to generate speech output from text, including concatenative synthesis, unit selection synthesis, diphone synthesis, domain- Including, but not limited to, specific synthesis, formant synthesis, articulatory synthesis, hidden Markov model (HMM) based synthesis, and sinusoidal synthesis.

In some examples, speech synthesis module 64 is configured to synthesize individual words based on phoneme strings that correspond to the words. For example, phoneme strings are associated with words in the generated text string. Phoneme strings are stored in metadata associated with words. The speech synthesis module 64 is configured to directly process phoneme strings in the metadata to synthesize words in the form of speech.

Because cloud environments generally have more processing power or resources than client devices, it is possible to obtain higher-quality speech output from client-side synthesis than in reality. However, the present invention is not limited to this, and of course, the voice synthesis process can actually be performed on the client side (see Figure 5).

Meanwhile, according to an embodiment of the present invention, the cloud environment may further include an intelligent agent (Artificial Intelligence Agent, AI agent) 62. The intelligent agent 62 may be designed to perform at least some of the functions performed by the above-described ASR module 61, NLU module 62, and/or TTS module 64. Additionally, the intelligent agent module 62 may contribute to performing independent functions of the ASR module 61, NLU module 62, and/or TTS module 64.

The intelligent agent module 62 can perform the above-described functions through deep learning. The deep learning involves a lot of research (representation) when there is data in a form that a computer can understand (for example, in the case of images, pixel information is expressed as a column vector, etc.) and applying this to learning. Research is underway on how to create better expression techniques and how to create models to learn them, and as a result of these efforts, deep neural networks (DNN) and convolutional deep neural networks (CNN) are being developed. ), various deep learning techniques such as Recurrent Boltzmann Machine (RNN), Restricted Boltzmann Machine (RBM), deep belief networks (DBN), and Deep Q-Network. They can be applied to fields such as computer vision, speech recognition, natural language processing, and voice/signal processing.

Currently, all major commercial voice recognition systems (MS Cortana, Skype Translator, Google Now, Apple Siri, etc.) are based on deep learning techniques.

In particular, the intelligent agent module 62 can perform various natural language processing processes, including machine translation, emotion analysis, and information retrieval, using a deep artificial neural network structure in the field of natural language processing. You can.

Meanwhile, the cloud environment may include a service manager 65 that can support the functions of the intelligent agent 62 by collecting various personalized information. Personalized information acquired through the service manager includes at least one data (calendar application, messaging service, music application usage, etc.) used by the client device 50 through a cloud environment, the client device 50 and/or the cloud. At least one sensing data collected by (60) (camera, microphone, temperature, humidity, gyro sensor, C-V2X, pulse, ambient light, iris scan, etc.), the client It may include off-device data that is not directly related to the device 50. For example, the personalized information may include maps, SMS, News, Music, Stock, Weather, and Wikipedia information.

For convenience of explanation, the intelligent agent 62 is expressed as a separate block to be distinguished from the ASR module 61, NLU module 63, and TTS module 64, but the intelligent agent 62 is divided into each module ( 61,62,64) may perform at least part or all of the functions.

Above, in FIG. 4, an example in which the intelligent agent 62 is implemented in a cloud environment due to computing operation, storage, and power constraints has been described, but the present invention is not limited to this.

For example, Figure 5 is the same as shown in Figure 4 except that the intelligent agent (AI agent) is included in the client device.

Figure 5 shows a schematic block diagram of a voice synthesis device in a voice synthesis system environment according to another embodiment of the present invention. The client device 70 and the cloud environment 80 shown in FIG. 5 may correspond to the client device 50 and the cloud environment 60 mentioned in FIG. 4 with only some differences in configuration and function. Accordingly, reference may be made to FIG. 4 for specific functions of the corresponding blocks.

Referring to Figure 5, the client device 70 includes a preprocessing module 51, a voice activation module 72, an ASR module 73, an intelligent agent 74, an NLU module 75, and a TTS module. (76) may be included. Additionally, the client device 50 may include an input module (at least one microphone) and at least one output module.

Additionally, the cloud environment may include Cloud Knowledge 80, which stores personalized information in the form of knowledge.

The function of each module shown in FIG. 5 may refer to FIG. 4. However, since the ASR module 73, NLU module 75, and TTS module 76 are included in the client device 70, communication with the cloud may not be necessary for voice processing processes such as voice recognition and voice synthesis. , As a result, immediate and real-time voice processing operations are possible.

Each module shown in FIGS. 4 and 5 is only an example to explain the voice processing process, and may have more or fewer modules than the modules shown in FIGS. 4 and 5. Additionally, it should be noted that two or more modules can be combined or have different modules or different arrangements of modules. The various modules shown in FIGS. 4 and 5 may be implemented as one or more signal processing and/or custom integrated circuits, hardware, software instructions for execution by one or more processors, firmware, or a combination thereof.

Figure 6 shows a schematic block diagram of an intelligent agent capable of implementing voice synthesis based on emotion information according to an embodiment of the present invention.

Referring to FIG. 6, the intelligent agent 74 can support interaction with the user (interactive operation) in addition to performing ASR operations, NLU operations, and TTS operations in the voice processing process explained through FIGS. 4 and 5. there is. Alternatively, the intelligent agent 74 uses context information to perform an operation where the NLU module 63 clarifies, supplements, or additionally defines the information contained in the text expressions received from the ASR module 61. You can contribute.

Here, the context information includes the client device user's preferences, hardware and/or software states of the client device, various sensor information collected before, during, or immediately after user input, and previous interactions between the intelligent agent and the user. may include fields (e.g., conversations), etc. Of course, the context information in this document is a feature that is dynamic and varies depending on time, location, content of conversation, and other factors.

The intelligent agent 74 may further include a context fusion and learning module 91, local knowledge 92, and dialog management 93.

The context fusion and learning module 91 can learn the user's intention based on at least one data. The at least one data may include at least one sensing data acquired from a client device or a cloud environment. In addition, the at least one data includes speaker identification, acoustic event detection, speaker personal information (gender and age detection), and voice activity detection (VAD). , may include emotion information (Emotion Classification).

The speaker identification may mean specifying the speaker from a group of conversations registered by voice. The speaker identification may include a process of identifying a pre-registered speaker or registering a new speaker. Acoustic event detection goes beyond voice recognition technology and can recognize the type of sound and the location of the sound by recognizing the sound itself. Voice activity detection (VAD) is a speech processing technique in which the presence or absence of human speech (voice) is detected in an audio signal that may include music, noise, or other sounds. According to one example, the intelligent agent 74 may check whether speech exists from the input audio signal. According to one example, the intelligent agent 74 may distinguish speech data and non-speech data using a deep neural network (DNN) model. Additionally, the intelligent agent 74 can perform an emotion classification operation on speech data using a deep neural network (DNN) model. According to the emotion classification operation, speech data can be classified into Anger, Boredom, Fear, Happiness, and Sadness.

The context fusion and learning module 91 may include a DNN model to perform the above-described operations, and may confirm the intent of the user input based on the DNN model and sensing information collected from a client device or cloud environment. .

Of course, the at least one piece of data is merely an example, and any data that can be referenced to confirm the user's intention during voice processing may be included. Of course, the at least one data can be obtained through the above-described DNN model.

Intelligent agent 74 may include local knowledge 92. The local knowledge 92 may include user data. The user data may include the user's preferences, user address, user's initial language setting, user's contact list, etc. According to one example, the intelligent agent 74 may supplement the information included in the user's voice input using the user's specific information to additionally define the user's intention. For example, in response to a user's request to “invite my friends to my birthday party,” intelligent agent 74 may ask the user to determine who the “friends” are and when and where the “birthday party” will be held. The local knowledge 92 can be used without requiring the user to provide more specific information.

The intelligent agent 74 may further include dialog management (Dialog Management) 93. The intelligent agent 74 can provide a dialog interface to enable voice conversation with the user. The dialog interface may refer to a process of outputting a response to a user's voice input through a display or speaker. Here, the final result output through the dialog interface may be based on the above-described ASR operation, NLU operation, and TTS operation.

Figure 7 is another block diagram of a voice synthesis device based on emotion information according to an embodiment of the present invention.

The voice synthesis device (TTS Device, 100) shown in FIG. 7 may include an audio output device 110 for outputting voice processed by the TTS device 100 or another device.

Figure 7 discloses a voice synthesis device (TTS Device, 100) for performing voice synthesis. One embodiment of the present invention may include computer-readable and computer-executable instructions that may be included in the TTS device 100. Figure 7 discloses a plurality of components included in the TTS device 100, but of course, components that are not disclosed may also be included in the TTS device 100.

Meanwhile, some components disclosed in the TTS device 100 are single components and may appear multiple times in one device. For example, the TTS device 100 may include a plurality of input devices 120, output devices 130, or a plurality of controllers/processors 140.

Multiple TTS devices may be applied to one voice synthesis device. In such a multi-device system, the TTS device may include different components to perform various aspects of speech synthesis processing. The TTS device 100 shown in FIG. 7 is exemplary and may be an independent device or may be implemented as a component of a larger device or system.

One embodiment of the present invention can be applied to a plurality of different devices and computer systems, such as general-purpose computing systems, server-client computing systems, telephone computing systems, laptop computers, portable terminals, PDAs, tablet computers, etc. there is. The TTS device 100 is used in automatic teller machines (ATMs), kiosks, global positioning systems (GPS), home appliances (e.g., refrigerators, ovens, washing machines, etc.), vehicles, and e-book readers. It may also be applied as a component of another device or system that provides voice recognition functions, such as readers.

Referring to FIG. 7, the TTS device 100 may include a voice output device 110 for outputting voice processed by the TTS device 100 or another device. The audio output device 110 may include a speaker, headphone, or other suitable component for propagating audio. The voice output device 110 may be integrated into the TTS device 100 or may be implemented separately from the TTS device 100.

The TTS device 100 may include an address/data bus 224 for transferring data between components of the TTS device 100. Each component within the TTS device 100 may be directly connected to other components through the bus 224. Meanwhile, each component within the TTS device 100 may be directly connected to the TTS module 170.

The TTS device 100 may include a control unit (processor) 140. The processor 208 may correspond to a CPU for processing data, computer-readable instructions for processing data, and memory for storing data and instructions. The memory 150 may include volatile RAM, non-volatile ROM, or other types of memory.

The TTS device 100 may include storage 160 for storing data and commands. Storage 160 may include magnetic storage, optical storage, solid-state storage types, etc.

TTS device 100 may be connected to removable or external memory (e.g., removable memory card, memory key drive, network storage, etc.) through input device 120 or output device 130.

Computer instructions to be processed by the processor 140 for operating the TTS device 100 and various components may be executed by the processor 140, memory 150, storage 160, and external devices. Alternatively, it may be stored in memory or storage included in the TTS module 170, which will be described later. Alternatively, all or part of the executable instructions may be embedded in hardware or firmware in addition to software. An embodiment of the invention may be implemented in various combinations of, for example, software, firmware, and/or hardware.

The TTS device 100 includes an input device 120 and an output device 130. For example, the input device 120 may include an audio output device 110 such as a microphone, touch input device, keyboard, mouse, stylus, or other input device. The output device 130 may include a display (visual display or tactile display), audio speaker, headphone, printer, or other output device. Input device 120 and/or output device 130 may also include an interface for connecting external peripherals, such as Universal Serial Bus (USB), FireWire, Thunderbolt, or other connection protocols. Input device 120 and/or output device 130 may also include network connections, such as an Ethernet port, modem, etc. Wireless communication devices such as radio frequency (RF), infrared, Bluetooth, wireless local area network (WLAN) (WiFi, etc.) or wireless communication devices such as 5G networks, Long Term Evolution (LTE) networks, WiMAN networks, and 3G networks A wireless network may include wireless devices. The TTS device 100 may include the Internet or a distributed computing environment through the input device 120 and/or output device 130.

The TTS device 100 may include a TTS module 170 for processing audio waveforms including voice into text data.

TTS module 170 may be connected to bus 224, input device 120, output device 130, audio output device 110, processor 140, and/or other components of TTS device 100. there is.

The source of textual data may be generated by internal components of the TTS device 100. Additionally, the source of the text data may be received from an input device such as a keyboard, or may be transmitted to the TTS device 100 through a network connection. The text may be in the form of a sentence containing text, numbers, and/or punctuation for conversion to speech by the TTS module 170. The input text may also include a special annotation for processing by the TTS module 170, and the special annotation may indicate how the specific text should be pronounced. Text data can be processed in real time or stored and processed later.

The TTS module 170 may include a front end 171, a speech synthesis engine 172, and a TTS storage unit 180. The preprocessor 171 may convert the input test data into a symbolic linguistic representation for processing by the speech synthesis engine 172. The speech synthesis engine 172 may convert the input text into speech by comparing annotated phonetic unit models with information stored in the TTS storage unit 180. The preprocessor 171 and the voice synthesis engine 172 may include an embedded internal processor or memory, or may use the processor 1400 and memory 150 included in the TTS device 100. Commands for operating the preprocessor 171 and the voice synthesis engine 172 may be included in the TTS module 170, the memory 150 and storage 160 of the TTS device 100, or an external device.

Text input to the TTS module 170 may be transmitted to the preprocessor 171 for processing. The preprocessor 1710 may include modules for performing text normalization, linguistic analysis, and linguistic prosody generation.

While performing a text normalization operation, the preprocessor 171 processes text input and generates standard text, converting numbers, abbreviations, and symbols to the same as written ones. .

While performing a language analysis operation, the preprocessor 171 may analyze the language of the normalized text and generate a series of phonetic units corresponding to the input text. This process may be called phonetic transcription. Phonetic units include symbolic representations of sound units that are ultimately combined and output by the TTS device 100 as speech. Various sound units can be used to segment text for speech synthesis. The TTS module 170 supports phonemes (individual sounds), half-phonemes, di-phones (the last half of one phoneme combined with the first half of an adjacent phoneme), and biphones (bi-phones). Speech can be processed based on phones (two consecutive sounds), syllables, words, phrases, sentences, or other units. Each word can be mapped to one or more phonetic units. Such mapping can be performed using a language dictionary stored in the TTS device 100.

The language analysis performed by the preprocessor 171 also involves checking different grammatical elements such as prefixes, suffixes, phrases, punctuation, and syntactic boundaries. It can be included. These grammatical components can be used by the TTS module 1700 to create a natural audio waveform output. The language dictionary may also include letter-to-sound rules and other tools that can be used to pronounce previously unidentified words or letter combinations that may be generated by TTS module 170. . In general, the more information contained in a language dictionary, the higher quality voice output can be guaranteed.

Based on the linguistic analysis, the preprocessor 171 can perform linguistic prosody generation, where phonetic units are annotated with prosodic characteristics indicating how the final sound unit should be pronounced in the final output speech. there is.

The prosodic characteristics may also be called acoustic features. While performing this step, the preprocessor 171 may consider any prosodic annotations accompanying text input and integrate them into the TTS module 170. Such acoustic features may include pitch, energy, duration, etc. Application of acoustic features may be based on prosodic models available to the TTS module 170. These prosody models represent how phonetic units should be pronounced in specific situations. For example, a prosody model can describe a phoneme's position in a syllable, a syllable's position in a word, and a word's position in a sentence or phrase. phrase), neighboring phonetic units, etc. can be considered. Like a language dictionary, the more prosodic information there is, the more high-quality speech output can be guaranteed.

The output of the preprocessor 171 may include a series of phonetic units annotated with prosodic characteristics. The output of the preprocessor 171 may be referred to as a symbolic linguistic representation. The symbolic language expression may be transmitted to speech synthesis engine 172. The voice synthesis engine 172 performs a process of converting speech into an audio waveform for output to the user through the audio output device 110. Speech synthesis engine 172 may be configured to convert input text into high quality, natural speech in an efficient manner. Such high-quality speech can be configured to sound as similar to a human speaker as possible.

The voice synthesis engine 172 may perform voice synthesis using at least one or more different methods.

The Unit Selection Engine 173 compares the recorded speech database with the symbolic linguistic representation generated by the preprocessor 171. The unit selection engine 173 matches the symbolic language representation with speech audio units in the speech database. Matching units are selected and the selected matching units can be connected together to form a speech output. Each unit contains only an audio waveform corresponding to a phonetic unit, such as a short, wav file of a specific sound, along with a description of the various acoustic properties associated with the .wav file (pitch, energy, etc.). Alternatively, the speech unit may contain other information, such as a word, sentence or phrase, and the location at which it appears in neighboring speech units.

The unit selection engine 173 can match input text using all information in the unit database to generate a natural waveform. The unit database may include multiple examples of speech units that provide the TTS device 100 with different options for linking the units into speech. One advantage of unit selection is that, depending on the size of the database, natural, natural voice output can be generated. Additionally, the larger the unit database, the more capable the TTS device 100 is of constructing a natural voice.

Meanwhile, for voice synthesis, there is a parameter synthesis method in addition to the unit selection synthesis described above. In parametric synthesis, synthesis parameters such as frequency, volume, and noise may be transformed by a parametric synthesis engine 175, a digital signal processor, or another audio generation device to generate an artificial voice waveform.

Parametric synthesis can use acoustic models and various statistical techniques to match the desired output speech parameters to symbolic linguistic representations. Parameter synthesis not only allows voice processing without a large database related to unit selection, but also allows accurate processing at high processing speeds. The unit selection synthesis method and the parameter synthesis method may be performed individually or in combination to generate voice audio output.

Parametric speech synthesis can be performed as follows. The TTS module 170 may include an acoustic model capable of converting symbolic linguistic representation into a synthetic acoustic waveform of text input based on audio signal manipulation. The acoustic model may include rules that can be used by the parameter synthesis engine 175 to assign specific audio waveform parameters to input speech units and/or prosodic annotations. You can. The above rules can be used to calculate a score indicating the likelihood that a particular audio output parameter (frequency, volume, etc.) corresponds to a portion of the input symbolic language expression from the preprocessor 171.

The parameter synthesis engine 175 may apply a plurality of techniques to match the speech to be synthesized with input speech units and/or prosodic annotations. One common technique uses a Hidden Markov Model (HMM), which can be used to determine the probability that an audio output should match a text input. HMM can be used to convert parameters of language and acoustic space into parameters to be used by a vocoder (digital voice encoder) to artificially synthesize the desired voice.

The TTS device 100 may include a voice unit database for use in unit selection.

The voice unit database may be stored in TTS storage 180, storage 160, or other storage configuration. The speech unit database may include recorded speech utterances. The speech utterance may be text corresponding to the content of the utterance. Additionally, the speech unit database may contain recorded speech (in the form of audio waveforms, feature vectors, or other formats) that takes up significant storage space in the TTS device 100. Unit samples in a speech unit database can be classified in a variety of ways, including by phonetic unit (phoneme, diphone, word, etc.), linguistic prosodic label, acoustic feature sequence, speaker identity, etc. A sample utterance can be used to create a mathematical model corresponding to the desired audio output for a particular speech unit.

When matching a symbolic language expression, speech synthesis engine 172 may select a unit within a speech unit database that most closely matches the input text (including both phonetic units and prosodic annotations). In general, the larger the voice unit database, the greater the number of selectable unit samples, enabling accurate speech output.

Audio waveforms including voice output from the TTS module 213 may be transmitted to the audio output device 110 for output to the user. Audio waveforms containing speech may be stored in a number of different formats, such as a series of feature vectors, uncompressed audio data, or compressed audio data. For example, audio output may be encoded and/or compressed by an encoder/decoder prior to transmission. The encoder/decoder can encode and decode audio data such as digitized audio data, feature vectors, etc. Also, of course, the encoder/decoder function may be located in a separate component or may be performed by the processor 140 or the TTS module 170.

Meanwhile, the TTS storage 180 can store other information for speech recognition.

The content of TTS storage 180 may be prepared for general TTS use, or may be customized to include sounds and words likely to be used in specific applications. For example, for TTS processing by a GPS device, TTS storage 180 may include custom voices specialized for location and navigation.

Also, for example, TTS storage 180 may be customized to the user based on personalized desired voice output. For example, a user may prefer that the output voice be of a certain gender, a certain accent, a certain speed, or a certain emotion (eg, a happy voice). The speech synthesis engine 172 may include a specialized database or model to account for such user preferences.

TTS device 100 may also be configured to perform TTS processing in multiple languages. For each language, TTS module 170 may include data, commands and/or components specifically configured to synthesize speech in the desired language.

To improve performance, the TTS module 213 can modify or update the contents of the TTS storage 180 based on feedback on the TTS processing results, so that the TTS module 170 has the ability to provide from a training corpus. The above can improve voice recognition.

As the processing capability of the TTS device 100 improves, voice output is possible by reflecting the emotional attributes of the input text. Alternatively, the TTS device 100 can output voice by reflecting the intention (emotion information) of the user who wrote the input text, even if the emotion attribute is not included in the input text.

The TTS system may integrate the various components mentioned above and other components when the model to be integrated into the TTS module that actually performs TTS processing is built. As an example, the TTS device 100 may insert emotional elements into the voice.

In order to output a voice to which emotional information is added, the TTS device 100 may include an emotion insertion module 177. The emotion insertion module 177 may be integrated into the TTS module 170 or as part of the preprocessor 171 or the voice synthesis engine 172. The emotion insertion module 177 enables voice synthesis based on emotion information to be implemented using metadata corresponding to emotion attributes. According to an embodiment of the present invention, the metadata may use a markup language, preferably SSML (Speech Synthesis Markup Language). The emotion information-based voice synthesis method using the SSML format will be described in detail below.

Figure 8 is a flowchart of a voice synthesis method based on emotion information according to an embodiment of the present invention.

The emotion information-based voice synthesis method according to an embodiment of the present invention can be implemented in the TTS device described with reference to FIGS. 1 to 7. Hereinafter, the emotion information-based voice synthesis method and the operation of the voice synthesis device for implementing the method according to an embodiment of the present invention will be described in detail with reference to the necessary drawings.

Meanwhile, as described above, it is assumed that the emotion information-based voice synthesis method according to an embodiment of the present invention is implemented through communication between two users (U1 and U2) through part of the communication system shown in FIG. 3. This explains. In addition, the text message sent by the sending user (U1) is received through the receiving device 14 of the receiving user (U2), and the receiving device 14 corresponds to the voice synthesis device 100 shown in FIG. 7. The description will be made assuming that the process of adding emotional information to the voice corresponding to the text message sent by the sending user U1 through the voice synthesis device 100 and performing the final voice output operation is described.

Referring to FIG. 8, the voice synthesis device 100 can receive a message (S100).

The voice synthesis device 100 may transmit the received message to the voice synthesis engine (S101).

Additionally, the voice synthesis device 100 may determine whether the received message is a message in which emotional information is set (S110).

The emotional information can be divided into at least one of neutral, love, happy, angry, sad, worry, or sorry and set in the message. there is. A description of the operation of the transmitting device that creates and transmits the message will be described in more detail later with reference to FIGS. 17A and 17B. The message is not limited to a message received from a transmitting device through a wireless communication unit, and of course may also include a message input through the input device 120 described in FIG. 7.

When first emotion information is set in the message (S110:YES), the voice synthesis device 100 may transmit first metadata corresponding to the first emotion information to the voice synthesis engine (S111). Here, the voice synthesis engine may be the TTS module 170 described in FIG. 7 or the voice synthesis engine 172 included in the TTS module 170. The voice synthesis engine causes the first emotion information to be added to the voice corresponding to the received message and output.

When the voice synthesis device 100 determines that emotional information is not set in the received message, it may generate second emotional information through an inference process from the received message. To this end, the voice synthesis device 100 may transmit the received message to the semantic analysis module and the context analysis module (S113).

Here, the second emotion information is to be distinguished from the above-described first emotion information, and the first emotion information is emotion information set along with message creation in the transmitting device that sent the message. And the second emotion information is defined as emotion information newly generated by analyzing the message received from the voice synthesis device 100 when the message transmitted from the transmitting device does not have any emotion information set.

The voice synthesis device 100 can analyze the meaning within the sentences of the received message (S120).

The message may consist of at least one sentence, and analysis of meaning within the sentence may be based on emotional elements included in one sentence from which emotions can be inferred. The emotional element that can infer the emotion may include at least one of a symbol, an emoticon, and a text (word). Accordingly, the semantic analysis within the sentence may determine that each sentence has different emotional information.

In addition, even if the text is the same, the intention (emotion) of the user who delivered the text may vary, so the voice synthesis device 100 can infer the emotion through semantic analysis within the sentence based on various emotional elements. .

Meanwhile, when a plurality of emotional elements are extracted within one sentence, the speech synthesis device 100 may combine the plurality of emotional elements to infer emotional information to be reflected in the single sentence. Additionally, when a plurality of emotional elements exist in one sentence, the speech synthesis device 100 may generate second emotional information by additionally adding a greater weight to the result of the context analysis.

Additionally, the voice synthesis device 100 may analyze the context between a plurality of sentences constituting the received message (S122).

The voice synthesis device 100 may generate the second emotional information based on the entire context of the message. Even if it is the same sentence, there may be cases where the meaning within the sentence cannot be analyzed, and even if it is the same sentence, the feelings of the message sender may change over time during the entire process of sending and receiving the message. In this case, it is necessary to generate second emotional information by considering the context of the entire message.

The speech synthesis device 100 may generate the second emotional information based on the semantic analysis result and/or the context analysis result within the sentence (S130).

The voice synthesis engine performs a voice synthesis operation so that the above-described first emotion information or the second emotion information is reflected and output in the process of outputting the voice corresponding to the received message (S140).

The first emotion information and second emotion information transmitted to the voice synthesis engine may be transmitted in metadata format. The metadata is a markup language and may be SSML (speech synthesis markup language). The expression described in the SSML format may include elements that display the aforementioned emotional elements. Emotion information transmitted to the voice synthesis engine will be described in more detail with reference to FIGS. 14 and 15.

As above, the speech synthesis device 100 generates second emotional information through semantic analysis and context analysis within the sentence when first emotional information is set in the received message, or when emotional information is not set in the received message, We looked at the process of performing voice synthesis operations based on emotional information.

Hereinafter, a more specific process for generating second emotional information will be described through FIGS. 9 to 13.

Figure 9 is a flowchart of a voice synthesis method based on emotion information for generating new emotion information when emotion information is not set in a received message according to an embodiment of the present invention. The emotion information-based voice synthesis method according to an embodiment of the present invention can be implemented in the TTS device described with reference to FIGS. 1 to 7.

In addition, the speech synthesis device 100 according to an embodiment of the present invention may further include a semantic analysis module, a context analysis module, and an emotion judgment module, and the semantic analysis module, the context analysis module, and the emotion judgment module are shown in FIG. 7 It may be integrated into the TTS device 202 or the TTS module 170 or the voice synthesis engine 172 or the emotion insertion module 177 shown in and is a processor that overall controls the operation of the voice synthesis device 100. Alternatively, it may be stored and provided as a software program in memory.

Referring to FIG. 9, the voice synthesis device 100 can receive a message from a transmitting device (S200). The voice synthesis device 100 may transmit the received message to the semantic analysis module and the context analysis module.

The semantic analysis module may calculate a first emotion vector based on emotional elements included in the message from which emotions can be inferred, and transmit the first emotion vector to the emotion determination module (S210).

The first emotion vector expresses emotional items such as neutral, love, happy, angry, and sad as a single vector, and can be defined as a weight assigned to each of the plurality of emotional elements (or emotional items or emotional attributes) described above. there is. Here, the sum of the weights assigned to each of the plurality of emotional elements may be normalized.

For example, Figure 10 is a diagram for explaining an emotion vector according to an embodiment of the present invention. Referring to FIG. 10, when the received message (M1) is “I love you,” the voice synthesis device 100 calculates a first emotion vector through a semantic analysis module, and the first emotion vector includes a plurality of emotional elements. (EA1, EA2, EA3, EA4, EA5) may be vectors with weights set for each. For example, in the first emotion vector, for the message I love you (M1), the weights of neutral (EA1), angry (EA4), and sad (EA5) are given as "0", and love (EA2) is given a weight of "0". "0.9", happiness (EA3), is an emotion vector containing emotional information created by assigning "0.1".

Meanwhile, FIG. 10 illustrates an example of generating an emotion vector based on the meaning of a word (or text) in a sentence. However, even if the semantic content of the text is the same, the emotion vector may be set differently.

Figure 11 shows another example of calculating a first emotion vector by analyzing the meaning of the message when emotion information is not set in the received message according to an embodiment of the present invention.

Referring to FIG. 11, the semantic content of the received message “Where are you” may be the same for all (a), (b), and (c), but the voice synthesis device 100 according to an embodiment of the present invention is ( In cases a), (b), and (c), different emotion vectors can be calculated.

For example, in case (a), the symbol used with the text (where are you) is a question mark, which is judged to be a neutral emotional element, and by assigning a weight of "1" to "neutral" among the plurality of emotional items, "Where are you?" The first emotion vector for "can be calculated. Also, in case (b), the symbol used with the text (where is it) is the question mark "?" and the special symbol "^^", and by assigning weights of 0.6 and 0.4 to "love" and "happiness", respectively, among the plurality of emotion items, the first emotion vector for "Where are you?^^" can be calculated. Also, in case (c), the symbol used with the text (where are you) is "--;;", a combination of special symbols, and among the multiple emotion items, "angry" and "sadness" are given weights of 0.8 and 0.2, respectively. By giving, the first emotion vector for “Where are you--;;” can be calculated. That is, even if the message has the same content, the voice synthesis device 100 can calculate different first emotion vectors by analyzing the meaning within the sentence.

Referring again to FIG. 9, the context analysis module may calculate a second emotion vector based on the entire context of the message and transmit the second emotion vector to the emotion determination module (S220).

Hereinafter, the process of calculating the second emotion vector through context analysis will be described with reference to FIG. 12.

Figure 12 shows an example of calculating a second emotion vector by analyzing the context of the message when emotion information is not set in the received message according to an embodiment of the present invention.

Referring to FIG. 12, a plurality of sentences may be included in both the received messages (a) and (b), and the second emotion vector may be set differently for the current sentence “Where are you?” (M22). For example, in case (a), the current sentence “Where are you?” If judged solely on its own, a weight of "1" may be given to the neutral emotion item, but if the previous sentence "I miss you" is taken into consideration through context analysis, the current sentence "Where are you?" is equivalent to the emotion item "love" or "happiness." A large weight may be given to . In this case, the case where all weight was given to “love” was explained. Also, in case (b), the current sentence “Where are you?” If judged solely on its own, a weight of "1" may be given to the neutral emotion item, but by considering the previous sentence "It's been a while since the appointment time" through context analysis, the current sentence "Where am I?" is equivalent to the emotion item "angry" or "sad." By assigning large weights of 0.8 and 0.2 to ", respectively, a second emotion vector can be calculated.

Referring again to FIG. 9, the first emotion vector calculated from the within-sentence semantic analysis module and the second emotion vector calculated from the context analysis module may be transmitted to the emotion judgment module, wherein the first emotion The second emotion information may be determined based on the vector and the second emotion vector (S230).

More specifically, the emotion determination module adds the first vector to which the first weight is assigned and the second emotion vector to which the second weight is assigned (S231). Additionally, the emotion determination module may select the emotion attribute with the largest value as the summation result as the second emotion information (S233).

Figure 13 shows an example of determining final emotion information based on the first emotion vector and the second emotion vector when emotion information is not set in the received message according to an embodiment of the present invention.

Referring to FIG. 13, the emotion determination module may determine second emotion information for the received message M3 according to Equation 1.

Here, EV is the emotion vector, Ws is the first weight given to the first emotion vector according to semantic analysis within the sentence, EVs is the first emotion vector, and Wc is the second emotion vector according to context analysis. It is the second weight assigned, and EVc is the second emotion vector. And, the sum of the first weight (Ws) and the second weight (Wc) is 1.

As a result of analyzing the meaning of the current sentence (Where are you?) of the received message (M3), the first emotion vectors (EVs) are assigned weights of 0.6 and 0.4 to the emotion vector items “love” and “happiness,” respectively. Additionally, as a result of context analysis, the second emotion vector (EVc) may be assigned a weight of 1.0 to the emotion vector item “love”.

Here, the emotion judgment module sets the first weight (Ws) and the second weight (Wc) according to the weight that each of the semantic analysis results and context analysis results within the sentence contribute to determining the emotional information of the entire message. It can be adjusted adaptively.

The first weight (Ws) and the second weight (Wc) may each be initially set to 0.5, but for example, when it is determined that there is continuity between a plurality of sentences constituting the message, as shown in FIG. 13, The second weight (Wc) may be increased more than the first weight (Ws). According to one example, the emotion judgment module may set the first weight (Ws) to 0.4 and the second weight (Wc) to 0.6. Accordingly, the first weight (Ws) may be applied to the first emotion vector (EVs) and the second weight (Wc) may be applied to the second emotion vector (EVc), thereby summing the two emotion vectors. there is. As a result of the summation, as shown in FIG. 13, vector values corresponding to each of the plurality of emotional items constituting the emotional vector (EV) can be calculated. For example, “neutral” 0.0, “love” 0.84, “happy” 0.016, “angry” 0.0, and “sad” 0.0.

The emotion judgment module may select the emotion item (“love”) with the largest value (0.84) among the calculated vector values as final emotion information (second emotion information).

Referring again to FIG. 9, the voice synthesis device 100 may generate the selected second emotion information in a second metadata format (SSML format) (S240) and transmit it to the voice synthesis engine (S250).

The voice synthesis engine can control the second emotional information to be reflected in the output voice. For example, the final emotion information is transmitted to the unit selection engine 173 or the parameter synthesis engine 175 through the emotion insertion module 177 of FIG. 7, and the final synthesized voice is output through the audio output device 110. You can.

Meanwhile, the emotion judgment module determines the emotional information when a message directly contains emotional expressions (for example, symbols that directly express emotions, such as emoticons, etc. in a sentence), and the result of semantic analysis within the sentence contributes. It is determined that the bar is large, and the first weight Ws may be increased.

Additionally, the voice synthesis device 100 may record and store the first weight and the second weight for each sender who sent the message. For example, in the case of emotional information used when a voice synthesis device outputs a message received from a first user as a voice, the first weight is reflected more than the second weight, and this pattern is repeated. In this case, when the voice synthesis device synthesizes the message received from the first user and outputs the voice synthesis, the initial setting value for the first weight may be set to be greater than the second weight.

Above, the process of synthesizing and outputting a voice by adding emotional information to a message received from a voice synthesis device has been explained. Hereinafter, through FIGS. 14 and 15, the process of processing voice data to enable voice synthesis by carrying emotional information in the voice synthesis device and transmitting it to the voice synthesis device will be described.

Figure 14 is a flowchart of a voice synthesis method based on emotion information according to an embodiment of the present invention. The embodiment disclosed in FIG. 14 may be implemented between a transmitting device and a voice synthesis device, or may be integrated into the voice synthesis device. Hereinafter, an example implemented in a voice synthesis device will be described in FIG. 14, but it should be noted that the present invention is not limited to this.

The emotion information-based voice synthesis method according to an embodiment of the present invention can be implemented in the TTS device described with reference to FIGS. 1 to 7. Additionally, the embodiment disclosed in FIG. 14 may be implemented based on the embodiment described in FIGS. 8 to 13 and may be implemented in combination with at least some of the embodiments described in FIGS. 8 to 13.

Referring to FIG. 14, the emotion information-based voice synthesis device according to an embodiment of the present invention can receive a message from a transmitting device (S300). The voice synthesis device can determine whether emotional information is set in the received message (S320). The process of receiving a message from a transmitting device is the same as described above.

When the voice synthesis device determines that emotional information is set in the received message, metadata corresponding to the emotional information may transmit metadata corresponding to the emotional information to the voice synthesis device along with the message (S310 ).

The metadata can be delivered in various ways. For example, the metadata may be a markup language such as Extensible Markup Language (XML) or Speech Synthesis Markup Language (SSML). The SSML is a markup language standard for synthesizing speech, and is disclosed through https://www.w3.org/TR/2010/REC-speech-synthesis11-20100907/ . The markup language may be composed of elements, and each element has an attribute.

The voice synthesis device according to an embodiment of the present invention can add an “emotion” element to the SSML standard to convey emotional information.

It can be described as <emotion="attribute">Sentence</emotion>, and the attribute values are neutral, love, happy, anger, sad, as described above. It can be expressed as emotional information such as (sadness), worry, or sorry. Additionally, the attribute value can be expanded into various forms.

For example, Figure 15 shows an example in which emotional information is described in SSML language format according to an embodiment of the present invention. Referring to FIG. 15, “xml version” is an element representing the xml version, “speak version” is an element representing the SSML version, and “s” is an element representing a sentence.

<emotion="happy">I'll wait.</emotion>

<emotion="love">I love you.</element>

As emotional information is added, the voice synthesis device can control voice synthesis to utter text such as “I’ll wait” with a happy emotion or “I love you” with a loving emotion.

According to an embodiment of the present invention, when emotional emotion information is set in a message received from a voice synthesis device, the set emotional information is processed into metadata in SSML format and transmitted to the voice synthesis engine. However, the transmitting device, Separately from the receiving device, a transmitting unit (transmitting device) that generates the metadata may be provided independently. However, in this case, if emotional information is not set in the message transmitted from the transmitting device, the function of simply transmitting the received message to the above-described voice synthesis device is performed, and the voice synthesis device performs the above-described process. 2 Speech synthesis can be performed by generating emotional information.

Figure 16 is a flowchart for explaining the operation of a voice synthesis system according to an embodiment of the present invention.

An emotional information-based voice synthesis system according to an embodiment of the present invention includes a transmitting device for transmitting a message, and a metadata corresponding to the emotional information extracted through a semantic analysis process and a context analysis process for the message received from the transmitting device. It may include a server that defines data in SSML (Speech Synthesis Markup Language) format and transmits it together with the message, and a voice synthesis device that synthesizes a voice in which the emotional information is reflected based on the metadata.

Referring to FIG. 16, the operating method of the voice synthesis system may include transmitting a message from a transmitting device (S400). Additionally, metadata corresponding to the emotional information received from the transmitting device can be defined in SSML format and transmitted to the voice synthesis device along with the message (S410). The voice synthesis engine may synthesize a voice reflecting emotional information based on the metadata (S420)

Here, the process of setting emotional information in a transmitting device and transmitting a message will be described in more detail with reference to FIGS. 17A and 17B.

Figures 17a and 17b show an example of sending a message by setting emotional information according to an embodiment of the present invention.

Referring to FIG. 17A, the transmitting device 12 can execute an application for creating a text message and receive a text message. In this case, an emotion setting menu (Es) may be provided. The emotion setting menu (Es) may be displayed as an icon. When the emotion setting menu (Es) is selected after a text message is input, the transmitting device 12 may display at least one emotion item on the display so that the user can select emotion information for the input text message. there is. According to one example, the at least one emotion item may be provided as a pop-up. According to one example, the at least one emotion item may be recommended as at least one candidate emotion item frequently used by the user.

Meanwhile, referring to FIG. 17B, according to one example, an emotion item may be set for each message (or each sentence). For example, after the first sentence is input, when a predetermined time has elapsed, the input message is displayed in the message display window, and the message with an emotion setting menu displayed may be provided. Accordingly, when a sentence is completed, the emotional information desired by the user can be set in the completed sentence itself.

Additionally, according to one example, after a specific emotion item is selected, the selected specific emotion item may be changed before the message is transmitted.

Additionally, according to one example, the transmitting device 12 may recommend an emotion item to be set with the message depending on the recipient. Alternatively, according to one example, the transmitting device 12 may learn the user's emotion item setting pattern and automatically set and display emotional information in consideration of the message content and the context of the previous sentence.

Meanwhile, the transmitting device 12 can also receive text messages through voice recognition. In this case, the transmitting device 12 can automatically set the user's emotional information from the input voice by analyzing the spectrum of the input voice. Alternatively, the transmitting device 12 may be equipped with a user voice analysis engine to learn the difference between the user's voice and emotions input in the past and automatically extract emotional information from the user's voice input based on the learned data. there is.

17A and 17B assume that when a text message is received from the transmitting device 12, a function for setting emotional information is provided in the message to be transmitted. However, the present invention is not limited to this. For example, if the above-described emotion information setting function does not exist in the transmitting device 12, emotion information-based voice synthesis can be implemented through the operation of generating the second emotion information disclosed through FIGS. 8 to 13. there is.

An emotional information-based voice synthesis device according to another aspect of the present invention includes a communication unit that receives a message, a voice synthesis unit that synthesizes a voice corresponding to the message, and a device that determines whether emotional information is set in the received message. It may include a processor that controls the voice synthesis unit to perform a voice synthesis operation based on emotional information.

When first emotion information is set in the message, the processor transmits first metadata corresponding to the first emotion information to the voice synthesis unit, and when emotion information is not set in the message, it is included in the message. Second emotion information may be generated based on at least one of an emotion element capable of inferring an existing emotion or the context of the message, and second metadata corresponding to the second emotion information may be transmitted to the voice synthesizer.

The voice synthesis unit synthesizes a voice corresponding to the message by adding emotional information determined based on either the first metadata or the second metadata.

A computing device according to another aspect of the present invention includes a processor and a memory including instructions executable by the processor, wherein the instructions determine whether emotional information is set in the received message.

The command is to transmit first metadata corresponding to the first emotion information to the speech synthesis engine when first emotion information is set in the message, and when emotion information is not set in the message, to the message. Second emotion information is generated based on at least one of the emotion elements from which emotions can be inferred or the context of the message, and second metadata corresponding to the second emotion information is transmitted to the speech synthesis engine.

The voice synthesis engine synthesizes a voice corresponding to the message by adding emotional information determined based on either the first metadata or the second metadata.

The voice synthesis method according to an embodiment of the present invention can be applied in various patterns. In other words, the voice synthesis method according to an embodiment of the present invention can be applied in various ways other than synthesizing voices by adding emotions to a received message.

For example, the voice synthesis method according to an embodiment of the present invention can be applied to audio books. Conventional audio books output synthesized voice content with the same tone, but when an embodiment of the present invention is applied, the meaning and context that the audio book is intended to convey can be analyzed to implement an audio book based on emotional information. .

In addition, for example, in the case of multi-media content (movies, dramas, animation dubbing, etc.), conventionally, lines were output through synthesized voices with the same tone, but when an embodiment of the present invention is applied, the lines and situations are As voices synthesized with various emotions can be output, the user's realistic content experience can be enriched.

In addition, for example, in the case of a navigation device, navigation was conventionally provided using a synthetic voice with the same tone, but when an embodiment of the present invention is applied, the driver's carelessness and warnings are avoided by speaking in various tones depending on the driving situation. The situation can be ventilated appropriately.

Additionally, for example, in the case of chatbots, conversations can be read in a voice with various emotions and styles appropriate for the situation during consultation.

The present invention described above can be implemented as computer-readable code on a program-recorded medium. Computer-readable media includes all types of recording devices that store data that can be read by a computer system. Examples of computer-readable media include HDD (Hard Disk Drive), SSD (Solid State Disk), SDD (Silicon Disk Drive), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. This also includes those implemented in the form of carrier waves (e.g., transmission via the Internet). Accordingly, the above detailed description should not be construed as restrictive in all respects and should be considered illustrative. The scope of the present invention should be determined by reasonable interpretation of the appended claims, and all changes within the equivalent scope of the present invention are included in the scope of the present invention.

Claims (15)

receiving a message;
determining whether emotional information is set in the received message;
When first emotion information is set in the message, transmitting first metadata corresponding to the first emotion information to a voice synthesis engine;
When emotion information is not set in the message, generate second emotion information based on semantic analysis and context analysis of the message, and transmit second metadata corresponding to the second emotion information to the speech synthesis engine. step; and
The voice synthesis engine includes the step of synthesizing a voice corresponding to the message by adding emotional information determined based on either the first metadata or the second metadata,
The step of generating the second emotional information is,
passing the message to a semantic analysis module and a context analysis module;
The semantic analysis module includes: transmitting a first emotion vector calculated based on emotional elements included in the message from which emotions can be inferred to an emotional information determination module;
The context analysis module includes transmitting a second emotion vector calculated based on the entire context of the message to the emotion information determination module;
Adding a first emotion vector to which a first weight is applied and a second emotion vector to which a second weight is applied; and
As a result of the summation, selecting an emotional attribute with the largest value as the second emotional information,
The first emotion vector is defined as a weight assigned to a plurality of emotional attributes by normalizing the sum of the weights,
The second emotion vector is a speech synthesis method based on emotion information, characterized in that the sum of the weights is normalized and defined as weights assigned to the plurality of emotional attributes. According to claim 1,
A voice synthesis method based on emotion information, characterized in that the first emotion information is set separately from the input of the message. According to claim 1,
The first metadata and the second metadata are described in a markup language format, and the markup language includes SSML (Speech Synthesis Markup Language). . According to claim 3,
The SSML includes elements representing emotional attributes,
The emotional attribute includes at least one of neutral, love, happy, angry, sad, worry, or sorry. Information-based speech synthesis method. delete According to claim 1,
The emotion information determination module may include determining the second emotion information to be reflected in voice synthesis based on the first emotion vector and the second emotion vector; and
generating the second metadata corresponding to the second emotion information and transmitting it to the speech synthesis engine;
A voice synthesis method based on emotional information, further comprising: delete delete According to claim 1,
increasing the first weight when the message contains an emotional expression; and
Further comprising: increasing the second weight when it is determined that there is continuity between the plurality of sentences constituting the message,
A voice synthesis method based on emotion information, characterized in that the first weight and the second weight are normalized and defined. According to clause 9,
The first weight and the second weight are recorded and stored for each sender of the message. According to claim 1,
The weights assigned to the plurality of emotional attributes constituting the first emotional vector are:
A speech synthesis method based on emotional information, characterized in that the result of inferring the semantic contents included in the message is given in consideration of symbols or graphic objects included in the message. According to claim 1,
The weights assigned to the plurality of emotional attributes constituting the second emotional vector are:
A speech synthesis method based on emotional information, characterized in that it is given by considering the context between sentences from which the flow of context can be inferred. According to claim 1,
The emotional elements that can infer the above emotions are:
A voice synthesis method based on emotion information, characterized in that it is defined as at least one of a text, symbol, or graphic object included in the received message. delete A communication unit that receives messages;
a voice synthesis unit that synthesizes voice corresponding to the message; and
A processor that controls the voice synthesis unit to perform a voice synthesis operation based on emotional information based on whether emotional information is set in the received message.
The processor,
When first emotion information is set in the message, first metadata corresponding to the first emotion information is transmitted to the voice synthesizer,
When emotional information is not set in the message, second emotional information is generated based on at least one of emotional elements that can infer the emotion included in the message or the context of the message, and the second emotional information Passing second metadata corresponding to the voice synthesis unit,
The voice synthesis unit,
Synthesizing a voice corresponding to the message by adding emotional information determined based on either the first metadata or the second metadata,
When the processor generates the second emotion information,
Passing the message to the semantic analysis module and the context analysis module,
The semantic analysis module transmits the first emotion vector calculated based on emotional elements included in the message and capable of inferring emotion to the emotion information determination module,
The context analysis module transmits a second emotion vector calculated based on the entire context of the message to the emotion information determination module,
Adding the first emotion vector to which the first weight is assigned and the second emotion vector to which the second weight is assigned, and
As a result of the summation, the emotional attribute with the largest value is selected as the second emotional information,
The first emotion vector is defined as a weight assigned to a plurality of emotional attributes by normalizing the sum of the weights,
The second emotion vector is a speech synthesis device based on emotion information, characterized in that the sum of the weights is normalized and defined as weights assigned to the plurality of emotional attributes.

Images